Intel adds 22nm octo-core 'Haswell' to CPU design roadmap
Intel's 'Tick, Tock' model of chip development is now well established and mapped out through the transition from 45nm to 32nm. Now we know what's happening at the 22nm.
The 22nm era, for Intel, runs through 2011 and 2012. The first of those years will see the arrival of 'Ivy Bridge', the 22nm die-shrink of the brand new, 32nm …
This topic is closed for new posts.
Posted Tuesday 19th August 2008 10:02 GMT
amanfromMars
Empty Headed Late Promises
#

I don't understand, Tony, why Intel are still pratting about and why they just don't go the whole Boss Hog Nine Yards and moot the Roswell Chipset for Future Processing Needs and Feeds, for that is surely where Rock and Victoria Falls is at in the Opposition.
The Only Viable Explanation is that they do not have the Future Technology. I suppose this afternoon will tell the Tale. I look forward to your ramblings on San Francisco later, to hear if they are anywhere near close to the Speed on ITs Leading Architecture.
Shalom for Now and Sayonara for Later and those are not Freudian Slips.
Posted Tuesday 19th August 2008 10:44 GMT
druck
Say it with pictures
#
Wouldn't some sort of graph be clearer, instead of all these names and numbers? Pretty much like the chip makers do themselves, with road maps showing how the architecture lines evolve.
Posted Tuesday 19th August 2008 10:44 GMT
pctechxp
CPU treats
#
I wonder how quick these are going to be and what treats await on the gaming front.
Posted Tuesday 19th August 2008 10:44 GMT
Mark Fenton
Bl**dy Hell! That was quick!
#

Had a quick google on "Fused Multiply Add" and looked at the Wikipedia page - and it referenced this article!
So either you guys are updating wiki *before* telling us stuff (shame on you! We should come first! :) ) or we have some real avid Wiki updaters on here....
Posted Tuesday 19th August 2008 12:33 GMT
Simon Ward
Nice to know ...
#

... that I'm not the only one who read 'Haswell' as 'Roswell' :-)
It's the one with the surgical kit in the pocket, ta.
Posted Tuesday 19th August 2008 12:33 GMT
alyn
More logical
#
Wouldn't 9 cores be more logical, arranged 3x3 on the wafer. 16 cores is an obvious next progression
Posted Tuesday 19th August 2008 12:33 GMT
Uwe Dippel
@pctechxp
#

pctechxp,
gaming isn't limited by CPU-power these days. Just FYI, and only if you didn't know it yet: Gaming is a problem of the graphics, so the graphics card is one bottleneck, the second being the cycles that you get from the CPU. Unfortunately, our current games cannot make use of more than 2 cores, and even those are used in a very unsymmetrical manner: 1 core running the operating system, and the other core runs the game. And that's no problem of operating system or CPU, but one of the game coders. If one wanted to use all cores, the games would need to be written in a different manner.
Paris, because she knows how to play
Posted Tuesday 19th August 2008 13:38 GMT
Eddie
@More Logical
#

Or a 2x2x2 cube?
Posted Tuesday 19th August 2008 14:51 GMT
Charles
@alyn
#
Actually, if you use a 3x3 logic grid, then eight cores would be the ideal number--the ninth (probably center) space would be given over to the hardware that's common to all eight cores, such as control hardware and common (say L3) cache.
Posted Tuesday 19th August 2008 16:52 GMT
Anonymous Coward
whoa there uwe
#

"Unfortunately, our current games cannot make use of more than 2 cores, and even those are used in a very unsymmetrical manner: 1 core running the operating system, and the other core runs the game."
That's just a load of stupid.
Paris, because she might know something of that.
Posted Tuesday 19th August 2008 16:52 GMT
David Halko
Fuzzy Roadmap, SUN has had 8 core for years.
#

Figuring the features and roadmap was somewhat difficult from the description.
When did Intel go 64 bit? 1 year after AMD went 64 bit with their proprietary x86 processor? 9 Years after SUN and other Open SPARC vendors went 64 bit?
Let me get this straight... Intel will release a proprietary x86 8 core CPU approximately 2012? Is this 6 years after SUN released an 8 core Open Sourced SPARC CPU?
Posted Tuesday 19th August 2008 16:57 GMT
Anonymous Coward
@alyn:More Logical
#

In terms of space yes 3x3 would fit nicely but remember how these chips work and how they access memory. If chip 1 wants to talk to chip 9 (or 3 to 7 etc) its a long path compared to 1 to 2, 1 to 4 etc. And they all have to talk to the memory controller and the hardware that co-ordinates them.
The best design sees the 3x3 arrangement with the centre chip poped out and in goes the central control (mem controller, synchro kit etc) and bingo 8 cores.
so you get this progression of 1 core, 2 cores (side by side) 4 cores (quadrents) 8 cores (3x3).
Posted Tuesday 19th August 2008 20:46 GMT
David Gillies
FMA? In 2012?
#

It's pretty pathetic that they've waited this long to add FMA to the architecture. The PowerPC 604e in my old PowerMac 8500 had that back in the mid 90's. I wrote some bitching matrix multiplication code in assembler that used it.
Posted Tuesday 19th August 2008 20:46 GMT
Bounty
where's the glued together version
#
Where is the 4 core glued to 4 core on the same package version of the 8 core?
Posted Wednesday 20th August 2008 05:26 GMT
alex d
The end is nigh!
#
22nm in three years. That's crazy!
You guys all know that the END comes at 16nm, right? No molecular computer could be any smaller...
Posted Wednesday 20th August 2008 19:29 GMT
Doug Lynn
AMD is going next to 6 core discrete and 12core dual 6 core
#
HI, as usual Intel is behind is the innovative department...
Posted Wednesday 20th August 2008 21:05 GMT
Craig Gorsuch
The end is not *quite* nigh
#

If a nano meter is 1x10^-9 meters and the physical limits of quantum mechanics is 1x10^-43 meters, I'd say we've got a way to go. However these dimensions are at least 1x10^20 times smaller than electrons. (If you didn't know, at 1x10^-43 matter becomes everywhere at once. Ask a quantum physicist.)
Still... There are limits to how small we can get and still maintain structural integrity with those high-speed electrons slamming into the walls of the processor pathways.
Don't know when we'll "hit the limit", and be forced to measure processor performance in a completely new way. We're obviously already moving past clock speed and adopting die-size. What's the next point? Running the "number-of-simultaneous-threads-in-a-single-instruction-set" root of the FSB multiplied by interconnect length divided by pipeline width?
Posted Sunday 24th August 2008 09:51 GMT
THOMAS STEWART VON DRASHEK
64 bit Media Center OK.
#

Just got MSN tip to download msn olympics player for ultimate media center. works ok & has special 64 bit edition. that with 64 bit silverlight should make 64 bit practical, beyond FLASH problems.maybe adobe gets better pitch from microsoft? anyway low gb memory & few remaining IE player probs seem likely to quickly be fading from ultimate 64, so bring on CPU & other good stuff. aHSO, Hyper Visor has been describe as E-Z & straightforward, so it should all click together.
Signed:PHYSICIAN THOMAS STEWART VON DRASHEK M.D.
This topic is closed for new posts.